The world is just not quite ready for humanoids yet (techcrunch.com)

Robotics pioneer Rodney Brooks warned that a surge of venture funding into humanoid startups risks creating a bubble because current robots can’t learn the fine motor dexterity needed for most real-world tasks. His critique—echoed by robotics VCs and AI leaders at Nvidia—says demos and valuations (e.g., Figure’s $39B raise) overstate practical readiness. Concerns center on safety in shared human spaces, cybersecurity, unclear timelines for commercialization, and weak unit economics if robots remain costly, fragile, or teleoperated rather than autonomous. Technically, the gap is large: humanoids involve 60+ degrees of freedom and complex contact-rich manipulation problems that far exceed the more tractable 6-DOF systems; many high-profile demos have hidden human control. Even major players like Tesla have missed timelines for Optimus, and skeptics expect broad adoption to take years or more than a decade. That said, research and niche markets are progressing—companies like Proception and Loomia work on tactile sensing and dexterity, while K-Scale Labs and Hugging Face have sold preorders for simpler or desktop humanoids—suggesting interim products (wheeled or semi-humanoid designs, improved touch sensors) may emerge before fully capable humanoid assistants become practical.